Google has launched Lyria 3, its latest generative music model from DeepMind, in the Gemini app in beta. Users can create 30-second tracks — including lyrics, cover art, and custom styles — from text prompts or uploaded photos and videos, with the model also rolling out to YouTube's Dream Track for Shorts creators in additional countries. All generated audio is embedded with SynthID watermarks, and Google says Lyria 3 was trained with copyright and partner agreements in mind, with filters in place to check outputs against existing content.

A new way to express yourself: Gemini can now create music - Google Blog